cs.LG（2025-05-05）

📊 共 22 篇论文 | 🔗 1 篇有代码

🎯 兴趣领域导航

支柱九：具身大模型 (Embodied Foundation Models) (14) 支柱二：RL算法与架构 (RL & Architecture) (6 🔗1) 支柱一：机器人控制 (Robot Control) (1) 支柱八：物理动画 (Physics-based Animation) (1)

🔬 支柱九：具身大模型 (Embodied Foundation Models) (14 篇)

#题目一句话要点标签🔗⭐
1 SEFE: Superficial and Essential Forgetting Eliminator for Multimodal Continual Instruction Tuning 提出SEFE，通过消除表面和本质遗忘提升多模态持续指令调优性能 large language model multimodal
2 Radio: Rate-Distortion Optimization for Large Language Model Compression 提出基于率失真优化的Radio量化方法，用于大规模语言模型压缩。 large language model
3 HSplitLoRA: A Heterogeneous Split Parameter-Efficient Fine-Tuning Framework for Large Language Models HSplitLoRA：异构拆分参数高效微调框架，用于大规模语言模型 large language model
4 Knowledge Graphs for Enhancing Large Language Models in Entity Disambiguation 利用知识图谱增强大语言模型在实体消歧中的表现 large language model
5 A Note on Statistically Accurate Tabular Data Generation Using Large Language Models 提出概率驱动的提示方法，利用大语言模型更准确地生成表格数据 large language model
6 Enhancing Chemical Reaction and Retrosynthesis Prediction with Large Language Model and Dual-task Learning ChemDual：利用大语言模型和双任务学习提升化学反应和逆合成预测 large language model
7 EntroLLM: Entropy Encoded Weight Compression for Efficient Large Language Model Inference on Edge Devices EntroLLM：面向边缘设备，通过熵编码压缩LLM权重以实现高效推理。 large language model
8 Towards Cross-Modality Modeling for Time Series Analytics: A Survey in the LLM Era 综述：面向时间序列分析的跨模态建模，聚焦LLM时代的方法与应用 large language model multimodal
9 RetroInfer: A Vector-Storage Approach for Scalable Long-Context LLM Inference RetroInfer：一种向量存储方法，用于可扩展的长上下文LLM推理。 large language model
10 Towards Quantifying the Hessian Structure of Neural Networks 揭示神经网络Hessian矩阵近块对角结构的成因：架构与训练的双重作用 large language model
11 When Your Own Output Becomes Your Training Data: Noise-to-Meaning Loops and a Formal RSI Trigger 提出Noise-to-Meaning递归自提升模型(N2M-RSI)，揭示AI自反馈学习中复杂性增长的机制。 large language model
12 Less is More: Efficient Weight Farcasting with 1-Layer Neural Network 提出基于单层神经网络的权重远距离预测方法，提升大模型训练效率 large language model
13 Unlearning vs. Obfuscation: Are We Truly Removing Knowledge? 区分遗忘与混淆：提出DF-MCQ方法，实现LLM的真实知识移除与拒绝行为。 large language model
14 Rewriting Pre-Training Data Boosts LLM Performance in Math and Code 重写预训练数据提升LLM在数学和代码领域的性能 large language model

🔬 支柱二：RL算法与架构 (RL & Architecture) (6 篇)

#题目一句话要点标签🔗⭐
15 Optimizing Chain-of-Thought Reasoners via Gradient Variance Minimization in Rejection Sampling and RL 提出GVM-RAFT，通过梯度方差最小化优化思维链推理，显著提升数学推理性能。 reinforcement learning large language model chain-of-thought ✅
16 Transfer learning-enhanced deep reinforcement learning for aerodynamic airfoil optimisation subject to structural constraints 提出基于迁移学习的深度强化学习方法，用于满足结构约束的气动翼型优化 reinforcement learning deep reinforcement learning DRL
17 Graph Neural Network-Based Reinforcement Learning for Controlling Biological Networks - the GATTACA Framework 提出GATTACA框架，利用图神经网络强化学习控制生物网络，实现细胞重编程。 reinforcement learning deep reinforcement learning DRL
18 GeoERM: Geometry-Aware Multi-Task Representation Learning on Riemannian Manifolds 提出GeoERM，解决异构多任务学习中表示的非欧几何问题。 representation learning
19 T2S: High-resolution Time Series Generation with Text-to-Series Diffusion Models 提出T2S：基于扩散模型的高分辨率文本到时间序列生成框架，解决长度限制和泛化性问题。 flow matching multimodal
20 Optimizing LLMs for Resource-Constrained Environments: A Survey of Model Compression Techniques 综述：针对资源受限环境优化LLM的模型压缩技术 distillation large language model

🔬 支柱一：机器人控制 (Robot Control) (1 篇)

#题目一句话要点标签🔗⭐
21 A New Perspective To Understanding Multi-resolution Hash Encoding For Neural Fields 提出域操纵视角，解析Instant-NGP多分辨率哈希编码提升神经场性能的原理。 manipulation

🔬 支柱八：物理动画 (Physics-based Animation) (1 篇)

#题目一句话要点标签🔗⭐
22 Enabling Local Neural Operators to perform Equation-Free System-Level Analysis 提出局部神经算子框架，实现无方程系统级分析，加速PDE系统分析。 spatiotemporal

⬅️ 返回 cs.LG 首页 · 🏠 返回主页